Sora: 9 Awesome AI-generated Videos Made By Sora

Recently, OpenAI unveiled Sora, a text-to-video model which creates AI-generated videos from written prompts.

Sora can create complex scenes featuring multiple characters and motion, as well as understanding how elements from user prompts exist within Sora’s physical universe.

At present, Sora is only available to a select few Red Teamers; yet its ability to produce highly realistic images has attracted considerable buzz.

As such, we decided to showcase nine examples of Sora in action – nine AI-generated videos created using Sora.

This includes an evaluation of each video’s successes (and failures) as well as, whenever possible, a brief outline of its original prompt.

9 of the Best AI-Generated Videos Made with Sora

1. A Dog Learns to Type

Justine Moore, one of a16z’s investment partners, shared an entertaining example that is making waves: an AI-generated video showing a dog typing away on an electronic keyboard.

This video stood out due to its very realistic appearance, more closely reflecting something you might take with your smartphone and share via social media than something like Lady in Red.

Lighting and motion of the dog’s head appear realistic in this video; all that reveals that it was created using artificial intelligence are unusual movements of his paws (and, yes, his typing fingers!).

2. Snowy Tokyo City

This example, created and released by OpenAI as well, depicts pedestrians strolling through Tokyo City covered with snow.

The dynamic camera movement that follows the central couple in a scene, combined with snowfall, traffic in the background and pedestrians gives this video clip an authentic, three dimensional feel.

We recognized some clues that this video was produced using artificial intelligence: cars vanished as they crossed beneath the treeline and distortion of pedestrians to the right of the screen were an easy telltale.

Original Prompt

“Takumi City in Tokyo, snowed over by a beautiful snowfall, is bustling with activity. As the camera moves along city streets and follows various people enjoying this perfect weather for shopping at nearby stalls – where beautiful Sakura petals dance through the air along with snowflakes!”

3. The Lady in Red

OpenAI’s promo video featuring a woman wearing a leather jacket and red dress walking through Tokyo streets may be one of the most well-known artificial intelligence (AI)-generated videos ever produced with Sora.

In this video she can be seen walking in an urban environment filled with pedestrians. Every detail, even small ones like wet flooring, gives this clip an air of reality.

All we could make out was an odd blurring in the background and her uncharacteristic gait when walking.

Original Prompt

“A stylish woman walks casually down a Tokyo street adorned with neon signs and animated city signage. She wears a leather jacket in black, red dress with black boots, red lipstick and sunglasses – as well as her usual accessories – including red purse and sunglasses – to match. As she walks casually and confidentially across wet pavement reflected with colourful lights reflecting in damp surfaces; numerous pedestrians walk past her.”

4. A Cat Searching for Shelter

Tim Brooks, an OpenAI research scientist, posted a video featuring a cat exploring an alleyway.

This video caught our eye not only because of the realistic background filled with garbage, but also because the cat and its reflection in the floor looked so lifelike.

We did notice some odd motion in the cat’s walking and in its tail when moving in certain places, yet still felt this video accurately represented a cat looking for shelter (without rain as specified by prompt).

Original Prompt

“…a white and orange tabby alley cat can be seen racing along a back street alley in heavy rainfall, looking for shelter…”

5. Late Night Super car

Global Things posted an example on their X account showing a supercar driving through a city covered in rain-slickened streets.

At first glance, this video appeared authentic; its cityscape seemed well-populated and naturally lit.

Only issue we found with this video was its unnatural traffic flow on both sides. Two taxis suddenly appeared out of thin air near its end.

Original Prompt

“A super car driving through city streets was shot from behind. To our surprise, no rain fell despite its damp floor surface.”

Also Read: Pros & Cons of AI Coding Assistants

6. An Old Man Eating a Burger

Next is an Instagram clip shared by an AiBot account.

At first glance, the video appears real; both the man’s face and cheeseburger seem convincingly realistic; further adding depth and dimension is provided by people moving about in the background.

We noticed some unusual aspects in this clip. Lettuce and sesame seeds seemed out of place and when the man took a bite of his burger it did not have the expected result on its bun.

Original Prompt

“An old man with graying hair and glasses enjoys an exquisite burger made of fresh lettuce, sesame seed-flecked bun, golden-brown beef patties and cheese in a red coat dines at an apparent fast food restaurant.”

7. A Postapocalyptic Jellyfish

Joe Taylor, an OpenAI researcher, provided us with a stunning image in which we see a jellyfish moving across an imaginary post-apocalyptic landscape.

The jellyfish’s movements were very varied, featuring liquid aesthetics and tentacle movements characteristic of invertebrate marine organisms. We thought the backdrop cityscape added an authentic feel.

The primary issue with the video was that it appeared as though the jellyfish were stationary against a green screen instead of moving through their environment.

Original Prompt

“At dusk, a giant translucent jellyfish glides gracefully through an abandoned urban landscape on 35mm film.”

8. Puppies Playing in the Snow

One of OpenAI’s promotional videos caught our eye – featuring puppies playing in snow.

This photograph truly stands out for its intricate detail and lifelike movement of snowflakes, which instantly captured our interest.

Minor glitches were present in this video, such as snow not moving naturally in some spots and one of the puppies falling below the snowline during its later portion. But as machine learning and generative AI improve further, these inconsistencies should become less frequent.

9. Car Travels Multiple Landscapes

Bilawal SIDHU, creator of AI, demonstrated how his model transformed a video of a car driving through rural landscape into a new video in an interesting post on X by inputting it into his model and inputting its output back into it.

In the video, for example, the car can be seen driving along rural roads before changing to winter settings with dinosaurs or tropical jungle surroundings.

This video illustrates how users can add a unique flair to videos by changing their style.

The Bottom Line

Together, these videos show Sora has the potential to revolutionize AI-generated videos. While not yet publicly available, early videos demonstrate its capacity for producing highly immersive and realistic content.

Start using AI for video creation using free tools such as Runway Gen-2 Stable Video Diffusion and PixVers, taking care to be as specific as possible when setting your prompt so the model can produce what you’re after.

Also Read: Is Autoblogging.ai the Best AI Writing Tool? Examining its Strengths and Alternatives

FAQs

1. How can I access Sora?

Unfortunately, Sora is only currently available to red teamers who are actively testing its capabilities; most individuals must wait for its wider release which remains to be decided upon.

2. Does Sora AI are available to the general public?

Yes. Sora can only be accessed by an exclusive group of red teamers; OpenAI has not announced when Sora will become publicly accessible.

Nexus Article

Nexus Article
      Nexus Article logo

      Dive into a world of daily insights at Nexus Article. Our diverse blogs span a spectrum of topics, offering fresh perspectives to elevate your knowledge. Join us on this journey of exploration and discovery.

      Quick Links

      © 2024 Nexus Article All Rights Reserved.

      Nexus Article
      Logo